Characterizing Speech and Language Pathology Outcomes in Stroke Rehabilitation
Brooke Hatfield, MS, CCC-SLP Deborah Millet, MS, CCC-SLP Janice Coles, MS, CCC-SLP Julie Gassaway, MS, RN Brendan Conroy, MD Randall J. Smout, MS
The focus of our time together
Describe Practice-Based Evidence Clinical Practice Improvement (PBE-CPI) design Describe the Post Stroke Rehabilitation Outcomes Project (PSROP) Examine who receives SLP services in poststroke rehabilitation Examine how SLPs spend their time with patients in post-stroke rehabilitation
The focus of our time together
Review the results of early data analyses regarding outcomes Consider clinical implications of the findings Consider limitations and future data queries Try not to think about the mojitos we’re missing out on by being here
Practice-Based Evidence for Clinical Practice Improvement
Susan Horn, PhD
Institute for Clinical Outcomes Research 699 E. South Temple, Suite 100 Salt Lake City, UT 84102-1282 801 – 466- 5595 shorn@isisicor.com
PBE-CPI Study Design
Content and timing of individual steps of a health care process, to determine how to achieve: Superior medical outcomes Least necessary cost Across the continuum of care
Goals of PBE-CPI
Improve/standardize process factors Practitioner activities/interventions Other interventions Medications “what we did”
Goals of PBE-CPI
Control for patient factors Psychosocial/demographic Factors Disease(s) Severity of Disease(s) Physiologic signs and symptoms Multiple points in time “for whom we did it”
Goals of PBE-CPI
Measure outcomes Clinical Health status Length of stay Discharge location “how far we got” and “how much it cost to get there”
Characteristics of a PBE study
Non-experimental – follows outcomes of treatments actually prescribed Inclusive – uses pt. populations undergoing routine clinical care Pragmatic– uses actual clinical outcomes Lower cost
Advantages of PBE study
Can simultaneously study outcomes of a large variety of treatments Can ask complex questions regarding treatment sequence effects, conditional effectiveness Can look at treatment effectiveness in whole clinical populations
More heterogeneous – reflecting clinical reality Less pt. selection bias – no requirement to consent
PBE-CPI: an alternative to Randomized Control Trials
PBE – CPI Patient Variables Use severity of illness to measure comorbidities and disease severity All patient qualify RCT Patient Variables Eliminate patients who could bias results based on comorbidities and more serious disease 15% of patients qualify
PBE-CPI: an alternative to Randomized Control Trials
PBE-CPI Process Variables Methods for stabilization RCT Process Variables Treatment protocol
Measure all processes and use analysis findings to develop protocol associated with better outcomes
Specify explicitly every important element of the process of care for both treatment and control arms
PBE-CPI: an alternative to Randomized Control Trials
PBE-CPI Outcome Variables RCT Outcome Variables
Dynamic improvement based on combinations of interventions Result Effectiveness research (usual conditions)
Change based on one protocol
Result Efficacy research (ideal conditions)
PBE-CPI: an alternative to Randomized Control Trials
PBE-CPI Hypotheses many and vague Alternatives not discrete Local knowledge contributes
RCT Hypothesis clear
Alternatives discrete
Not depend on local knowledge
PBE-CPI: an alternative to Randomized Control Trials
PBE-CPI Confounders affect outcomes and are interesting Effects are large
RCT Confounders not interesting
Effects are small
Alternatives to RCTs in the Literature
Results from 2 NEJM studies “Average results of the observational studies were remarkably similar to those of the randomized, controlled trials”
New England Journal of Medicine 2000 (June 22, 2000) 342: 1878-92
Alternatives to RCTs in the Literature
Conclusions from JAMA study “Significant between-study variability was seen as frequently among RCTs as between RCTs and non-randomized studies.” “…may reflect differences in true treatment effects under different study settings and in different populations.”
JAMA (Aug 2001) 286, 7:821-830
So what is a PBE-CPI study?
Three-dimensional measurement framework
Patient variables, process variables, and outcomes
Adjusts for severity of illness Led by a transdisciplinary team that
Develops and frames questions Defines variables for study Gathers data Interprets findings Implements findings
Measuring and Adjusting for Severity of Illness
Comprehensive Severity Index (CSI®)
2,200 + individual criteria
5,500 disease specific groups
Disease-specific and overall severity levels on a scale of 0 – 4, continuous Fixed times for inpatient reviews
Admission Maximum Discharge review
Components of a PBE-CPI Study
Development of data collection tool to record practices for each point of contact Patients enrolled, normal practices recorded via point-of-care forms Retrospective chart review to record data regarding patient, processes, and outcomes Retrospective chart review using CSI® Data analyzed and findings interpreted
A Previous PBE-CPI Study
Nursing Home Study (NPLUS) 1996-1997
6 LTC provider organizations 109 facilities 2, 490 residents studied Outcomes
Developed pressure ulcer Healed pressure ulcer Hospitalization Systemic Infections
A Previous PBE-CPI Study
Outcome: Develop Pressure Ulcer
More likely to develop a pressure ulcer if:
The patient was male The patient was >85 years old Static pressure relieving devices were used Dependency in greater or equal to 7 ADLs Signs of dehydration Weight loss of more than 5% in last 30 days Mechanical devices such as catheters were used
A Previous PBE-CPI Study
Outcome: Develop Pressure Ulcer
Less likely to develop a pressure ulcer if:
Disposable briefs were used Toileting program was used Combination of newer SSRI and antipsychotic were used RN spent more than 15 minutes with the patient Fluid orders and nutritional supplements were used
Pressure Ulcer Clinical Outcomes Post Implementation
Development of pressure ulcers for highrisk residents decreased from
14% pre-implementation to 8.7% postimplementation is still decreasing
Post-Stroke Rehabilitation Outcomes Project (PSROP)
Using PBE-CPI to investigate the “black box” of post-stroke rehabilitation
6 US sites, 1 New Zealand A study team at each site:
Physician Nursing Social Work Psychology PT/OT/SLP Therapeutic Recreation
Post-Stroke Rehabilitation Outcomes Project (PSROP)
Facility
National Rehabilitation Hospital Univ. of Pennsylvania Med Center LDS Hospital Rehabilitation Ctr Legacy Health System
Location
Washington, DC Philadelphia, PA Salt Lake City, UT Portland, OR
Site Director(s)
B Conroy, MD R Zorowitz, MD D Ryser, MD F Wong, MD LA Sims, RN
Type
Freestanding Rehab Unit Rehab Unit Rehab Unit
Beds
128 24 26 33
Stanford University Hospital
Loma Linda Univ. Medical Center
Palo Alto, CA
Loma Linda, CA
J Teraoka, MD
M Brandstater, MD
Rehab Unit
Rehab Unit
17
40
Wellington/Kenepuru Hospitals
Wellington, NZ
H McNaughton, MD
Rehab Unit
25, 20
PSROP Study Questions
Which patient characteristics are associated with improved post-stroke outcomes? Which treatment interventions or combinations are associated with improved outcomes (when controlling for patient characteristics)? What is the optimal intensity and duration of various post-stroke treatment interventions?
PSROP Patient Variables
Age, gender, race Payer source Type and side of stroke Stroke location Admission FIM Case-mix group
Admission severity of illness Acute care hospital LOS Date/time of stroke symptom onset
PSROP Process Variables
Rehabilitation LOS Medications Nutritional process Pain management Time to first rehab Oxygen use
Specific therapy interventions Intensity, frequency, and duration of therapy interventions
PSROP Inclusion Criteria
Rehab diagnosis: ICD-9 code of 430-438.99, 997.02, or 852-853 18 years or older 1st inpatient rehab admission after current stroke If interrupted stay – remained in study if less than 30 day interruption No exclusion criteria or need for consent
PSROP Data Collection
Supplemental documentation from all rehab disciplines at point of contact Retrospective chart review for Auxiliary Data Module (records data re: all variables) and CSI® March 2001 – March 2003 200 patients/site for consecutive stroke admissions
PSROP Limitations
Reliance on chart review Limited acute care information No standard of initial stroke severity (e.g. NIH Stroke Scale) in all facilities CSI dependent on ICD-9 codes Point of care documentation
Train-the-trainer approach Burdensome to team members Paper – some may have been misplaced
The Depth of the Data Set
141, 511 point of care forms 235 variables captured once for each patient 71 complication/process variables that depended on each patient’s length of stay 15 variables captured daily over length of stay Up to 8 variables per medication administered
Point-of-Care Documentation All Clinical Services
Development of a taxonomy of rehab activities to capture in detail what clinicians do – it had to:
be quick (less than 1 minute) to complete record duration of activity and date of session break down into activities and interventions
Activities: general target area of the task – broad Interventions: strategies, cues, protocols, equipment, education, diagnostic tests
Point-of-Care Documentation SLP Activities
Pre-functional Swallowing Face/Neck mobility Speech/Intelligibility Voice Verbal Expression Alternative/Non-verbal Expression Written Expression
Auditory comprehension Reading comprehension Problem solving/reasoning Orientation Attention Memory Pragmatics Executive functional skills
Point-of-Care Documentation SLP Interventions
Adaptive and Compensatory Strategies Neuromuscular Interventions Modalities Devices Perception
Soft Tissue Work Education/Counseling Diagnostic Tests
Did not include standardized tests
Point-of-Care Documentation Additional Data Points
Date Time spent in:
Professional level of treating clinician
Co-treatment Formal assessment Supervisory/team input Preparation of activities Group treatment
SLP SLP-A SLP aide/tech SLP student
S p eech & L an gu age T h erap y R eh ab ilitation A ctivities
16315
Patient ID :
D ate of T herapy Session:
S a m p l e
T herapist:
/
T im e session begins:
/
:
INTERVENTION CODES Duration of Activity Adaptive & Compensatory Strategies: E nter in 5 m inute increm ents. 01. M em ory strategies 02. M otor speech strategies P re-F unctional A ctivity 03. Sw allow ing strategies 04. D iet m odification/evaluation Sw allow ing 05. A ttention/focus strategies 06. Point/gesture strategies 07. V isual strategies/cueing F ace/N eck M obility 08. V erbal strategies/cueing 09. A uditory strategies/cueing Speech/Intelligibility 10. T actile strategies/cueing 11. A nalysis & sum m ary strategies/cueing Neuromuscular Interventions: V oice 12. O ral m otor treatm ent/R O M 13. R espiratory treatm ents V erbal expression 14. V ocal treatm ents 15. R esonance treatm ents 16. NDT A lternative/non-verbal expression 17. D PN S 18. T herm al tactile stim ulation W ritten expression 19. Postural aw areness M odalities: A uditory com prehension 20. EM G 21. B iofeedback 22. E lectrical stim ulation R eading com prehension Devices: 23. Incentive spirom etry 24. M em ory book/aids P roblem solving/reasoning 25. Speaking valves 26. A ugm entative com m unication devices O rientation 27. C om puter 28. V isi pitch 29. N asal m anom eter A ttention Perception: 30. R ight or left side aw areness strategies Soft Tissue W ork: M em ory
31. 32. 33. Strengthening Stretching M yofascial release
Interventions E nter one intervention code per group of boxes.
15
06
P ragm atics
Um, we’re going to need a few more boxes
Fast forward to when
All of the charts are reviewed for CSI and ADM information All of the point-of-care forms are scanned Data was “cleaned”
There was a huge pile of information and data waiting to be organized!
Data Analysis
Manageable bites
For analysis, “blocks” of therapy time were identified 1 block of SLP treatment = 3 hours 1 block of OT treatment = 4 hours 1 block of PT treatment = 6 hours Allows for comparison of groups while eliminating natural recovery time
Grab your coffee, it’s time for statistics
Descriptive statistics
Study variables Used to compare patient, process, and outcome variables
Chi-tests
ANOVA
Continuous data Variables entered step-wise into model Importance determined via Wald chi-square
Logistic regression
General Outcomes of the PSROP US subjects: N = 1,161
Patient demographics
Male 51.8% White 61% Average age 66.0
Female Black
49.2% 23%
(18.6 – 95.5)
Common comorbidities
HTN 78.6% DM 30.8% CAD 22.5%
General Outcomes of the PSROP
Process demographics
Average time from onset to admission to rehab: 13.8 days Mean rehab LOS: 18.6
The stroke itself
Right 44.2% Left 42.5% Hemorrhagic 23.2% Ischemic 76.7%
Bilateral 10.5%
General Outcomes of the PSROP
Severity
CMG
Mild (101-103): 11.5% Moderate (104-107): 39.6% Severe (108-114): 42.5%
Admission FIM (mean)
Total: Motor: Cognitive:
61.0 40.1 21.0
Severity of illness per CSI (mean): 20.7 Discharge to home: 81%
Factors Associated with Outcomes
Interpreting outcomes
Associations vs. causations For patients who had __?__ on admission, they were more/less likely to achieve the outcome. For patients who spent more time in __?__, they were more/less likely to achieve the outcome.
Will need predictive validity studies to follow-up
Outcome: Discharge Motor FIM Moderate stroke
General Assessment
Age Female PT Interventions Formal assessment Bed mobility OT Interventions Toileting Transfers SLP Interventions Speech intelligibility
Auditory comprehension Voice
Problem solving
Brainstem/ Cereb
Mod motor imp. Admission motor FIM
Transfer
Gait
Home management
Upper extremity control
Advanced gait
Medications
General Interventions
Anti-Parkinson
Days onset to rehab LOS Opioid analgesics Atypical antipsychotics
Outcome: Discharge Motor FIM Moderate stroke– 1st tx block only
General
Assessment Age PT Interventions Sitting OT Interventions Bathing SLP Interventions Auditory comprehension
Female
Brainstem/ Cereb Mod motor imp.
Transfer
Gait
Feeding/Eating
Voice
Admission motor FIM
General Interventions Days onset to rehab Medications Muscle relaxant Opioid analgesics
LOS
Old anticonvulsants
Outcome: Discharge Cognitive FIM Moderate stroke
General Assessment Aphasia Admission cognitive FIM PT Interventions OT Interventions Feeding/Eating SLP Interventions Auditory comprehension Problem solving
General
Interventions
Medications
Anti-Parkinsons Opioid analgesics New SSRIs
Outcome: Discharge Cognitive FIM Moderate stroke– 1st tx block only
General Assessment Aphasia Age Admission cognitive FIM PT Interventions OT Interventions Toileting SLP Interventions Auditory comprehension
General Interventions
Medications Anti-Parkinsons Old Anti-nausea
Outcome: Discharge Motor FIM Severe stroke
General
Assessment
Age Black race Mild motor impairment Admission motor FIM Admission cognitive FIM General Interventions Days onset to rehab Enteral feeding
PT Interventions
Formal assessment Bed mobility Gait Advanced gait
OT Interventions
Home management
SLP Interventions
Swallowing Orientation Reading comprehension
Medications Anti-Parkinsons Modafinil Old SSRIs
Outcome: Discharge Motor FIM Severe stroke– 1st tx block only
General Assessment
Age Severe motor impairment No dysphagia
PT Interventions
Bed mobility Gait Advanced gait
OT Interventions
Home management
SLP Interventions
Admission motor FIM No dysphagia
Neurotropic meds
General Interventions
Days onset to rehab Enteral feeding
Medications
Other antidepressant Old SSRIs
LOS
Outcome: Discharge Cognitive FIM Severe stroke
General Assessment Aphasia Mood and cognitive disturbance Admission cognitive FIM Functional mobility Community integration General Verbal expression PT Interventions Advanced gait OT Interventions Grooming SLP Interventions
Auditory comprehension Orientation
Bed mobility
Problem solving
Interventions
LOS
Medications
Outcome: Discharge Cognitive FIM Severe stroke – 1st tx block only
General
Assessment Aphasia PT Interventions Advanced gait OT Interventions Bed mobility SLP Interventions Orientation Problem solving
Mood and cognitive disturbance
Race: other Max severity of illness Severe motor impairment Right brain stroke General Interventions
Medications
Admission cognitive FIM
Days onset to rehab LOS
Trends Across Disciplines
Patients involved in higher level activities in the first block of therapy despite their initial level of impairment were more likely to be “successful”
PT: early gait training OT: home management SLP: problem solving
What made the treating clinician decide to do these activities with severe patients? They had “nothing to lose”. Why didn’t they decide to do these activities? Staffing issues, size of the patient, time resources
Is there something to this?
The question: Does introduction of high-level SLP activities early in post-stroke rehabilitation correlate with improved outcomes for low to mid-level functioning communicators following stroke?
What we (thought we) knew about SLP practice in inpatient rehab
Assessment leads to identified deficit areas Treatment plans target deficit areas directly
lead to improvement in the deficit area
deficit in verbal expression + therapy targets verbal expression = measurable improvement in verbal expression
Interventions and activities are introduced in a hierarchy of complexity
•from simple to complex •more impaired patients start low in the hierarchy with simple tasks •more advanced patients start with more complex tasks because they can do them
PSROP data indicate there may be a more effective way…
How to address the question? ID a homogenous group Set an outcome Use logistic regression to ID the variables positively associated with achieving the outcome
Identifying a Homogenous Group
The patient subset
n=397
At least 1 documented SLP session
Over 90% of patients at 5 US sites
1 - 8 blocks (3 hour chunks) of SLP services Removed labeled aphasia
Concern over inaccurate recording More variability in treatment approaches for patients without aphasia
Removed patients at A FIM 6, 7 for Aud Comp and Verbal Expression
Identifying a Homogenous Group
Low-level communicators
Admission FIM 1 – 3 for Comprehension alone Admission FIM 1 – 3 Comprehension paired with FIM 1 – 3 Verbal Expression
Mid-level communicators
Admission FIM 4 -5 Comprehension alone Admission FIM 4-5 Comprehension combined with Admission FIM Verbal Expression 4 -5
Identifying a Measurable Outcome
Success
Change in FIM at discharge (D FIM): Verbal Expression and Auditory Comprehension
Low-level communicators
Expression: increase to > Level 4 Comprehension: increase by 2 levels
Mid-level communicators
Expression: increase to Level 6 or higher Comprehension: increase to Level 6 or higher
Classifying SLP Activities
Simple
Swallowing, speech intelligibility, voice, orientation, attention, pre-functional
Mid-level
Verbal expression, alternative/non-verbal expression, written expression, auditory comprehension, reading comprehension, memory, pragmatics
Cognitively-linguistically complex
Problem solving/reasoning, executive functioning skills
Amount and Timing of Treatment Provided
397 patients averaged:
16.4 SLP sessions 11.4 days 602 minutes
Patients per Block of SLP Treatment
Number of Patients
100 80 60 40 20 0 1 2 3 4 5 6 7 8 Blocks of Treatment
Who were these patients?
Short (1 block)
Mean age (years) 66.1
Medium (5 blocks)
56.7
Long (8 blocks)
67.3
Side of lesion (%) right
left bilateral
50.6
34.5 12.6
43.3
33.3 20.0
58.8
35.3 5.9
Site of lesion (%)
brainstem subcortical cortical 18.4 31.0 37.9 36.7 23.3 33.3 17.7 11.8 58.8
Who were these patients?
Short (1 block) Medium (5 blocks) Long (5 blocks)
Mean admission
Motor FIM 44.5 34.5 28.9
Cognitive FIM
CSI
19.9
18.5
15.6
24.0
17.4
20.9
Intensity of SLP services
Short (1 block) Medium (5 blocks) Long (8 blocks)
Mean LOS (days)
12.2
23.6
34.7
Days of SLP sessions SLP sessions during rehab SLP minutes during rehab
5.0 6.2 214
16.3 24.8 914
24.8 39.7 1439
% Time Across Short Stay (1 Block Total)
Problem solving 16% Alt/Nonverbal 0% Pragmatics 0% Written expression 3% Speech 7% Attention 5% Reading comprehension 6% Orientation 5% Verbal expression 12% Voice 2% PF/Not related 2% Executive functions 5% Swallowing 19%
Memory 9%
Auditory comprehension 9%
% Time in 1st block only; Short Stay Low level communicators (Comp 1 -3)
Alt/Non-verbal 1% Problem solving 13% Pragmatics 0% Executive functions 3% Swallowing 29%
Written expression 1% Reading comprehension 4%
Auditory comprehension 10% Verbal expression 10%
Orientation 9% Speech 3% Attention 4%
Memory Voice 0% 11%
PF/not related 2%
% Time in 1st block only; Short Stay Mid level communicators (Comp 4-5)
Executive functions 6% Alt/Non-verbal 0% Pragmatics 0% Written expression 3% Reading comprehension 6% Memory 8% Auditory comprehension 8% Problem solving 18% Swallowing 16%
Speech 11% Attention 6%
Voice 2% Orientation 1%
Verbal expression 14%
PF/Not related 1%
Complexity of activities (% time) for Short Stay (1 block)
45 40 35 30 25 20 15 10 5 0 Simple Mid Complex
Trends for a Short Stay
Averaged 6.2 SLP sessions over 12.2 day LOS Patients had more:
R CVAs Cortical lesions
Discharge: home Least amount of:
Cognitive and Motor FIM change Time in simple SLP activities; 1st block
% Time Across Medium Stay (5 blocks)
Problem solving 19% Executive functions 1% Swallowing 25%
Alt/Non-verbal 1%
Written expression 3%
Pragmatics 1% Memory 5% Reading comprehension 7% Auditory comprehension 9% Attention 7% Speech 6% Orientation 4% Verbal expression 10% Voice 2% PF/Not related 0%
% Time in 1st block only; Medium Stay Low level communicators (Comp 1-3)
Executive functions 2% Pragmatics 1% Alt/Non-verbal 1% Memory 2% Written expression 4% Reading comprehension 7% Verbal expression 8% Auditory comprehension 11% Orientation 5% Speech Attention 4% 3% Problem solving 13% Swallowing 39%
Voice 0% PF/Not related 0%
% Time in 1st block only; Medium Stay Mid level communicators (Comp 4-5)
Alt/Non-verbal 0% Pragmatics 1% Memory 3% Reading comprehension 4% Problem solving 14% Executive functions 2%
Swallowing 39%
Speech 7% Attention 6% Orientation 4% Voice 3%
Written expression 4% Verbal expression 7% PF/Not related 0%
Auditory comprehension 6%
Complexity of Activities for Short and Medium Stays
45 40 35 30 25 20 15 10 5 0 Simple Mid Complex 1 block 5 blocks
Trends for a Medium Stay
Averaged 24.8 SLP sessions over 23.6 day LOS Younger (56 years) Patients had more:
R CVAs Brainstem lesions Time in simple activities in 1st block
Swallowing
Time in auditory comp (in low level group)
Discharge: institution
% Time Across Long Stay (8 blocks)
Problem solving 20% Alt/Nonverbal 0% Written expression 4% Pragmatics 0% Memory 4% Attention 11% Speech 7% Orientation 4% Voice 2% Executive functions 2% Swallowing 25%
Auditory comprehension 5% Reading comprehension 8% Verbal expression 8% Not related 0%
% Time in 1st block only; Long Stay Low level communicators (Comp 1-3)
Written Pragmatics expression 0% Reading 1% comprehension Alt/Non-verbal 3% 1% Memory 5% Auditory comprehension 7% Verbal expression 11% PF/Not related 0% Voice 0% Orientation 7% Speech 6% Attention 7% Problem solving 5% Executive functions 0%
Swallowing 47%
% Time in 1st block only; Long Stay Mid level communicators (Comp 4-5)
Pragmatics 0% Alt/Non-verbal 0% Written expression 3% Reading comp 10% Memory 1% Verbal expression Auditory comp 0% 4% Voice PF/Not related 4% 0% Speech 4% Orientation 2% Attention 19% Problem solving 18% Executive functions 0% Swallowing 35%
Complexity of Activities for Short, Medium, and Long Stays
60 50 40 30 20 10 0 Simple Mid Complex 1 block 5 blocks 8 blocks
Trends for a Long Stay
Averaged 34.7 SLP sessions over 39.7 day LOS Patients had more: R CVAs Cortical lesions Time in swallowing: 47% of 1st block Time in attention (mid level group) Almost no time in executive function across LOS Discharge: home
Activities in the 1st block of therapy
70 60 50 40 30 20 10 0 Simple Mid Complex short stay, low short stay, mid medium stay, low medium stay, mid long stay, low long stay, mid
How did these patients do?
Short (1 block) Medium (5 blocks) Long (8 blocks)
Mean increase motor FIM
cognitive FIM
19.6 4.6
23.1 5.3
30.1 8.1
Discharge destination (%)
home/community institution
81.6 18.4
60.0 40.0
88.2 11.8
Factors Associated with Success
A reminder of “success”:
Low-level communicators
Expression: increase in D FIM to > Level 4 Comprehension: increase in D FIM by 2 levels
Mid-level communicators
Comprehension: increase D FIM to > Level 6 Expression: increase D FIM to > Level 6
If the patient’s comprehension was 1 – 3 on admission
they were more likely to be successful if
in the first 3 hours of SLP they spent time in:
Problem solving Executive function
and/or they:
stayed longer
If the patient’s comprehension was 1 – 3 on admission
they were less likely to be successful if in the first 3 hours of SLP they spent time in:
Verbal expression Written expression
and/or they: were female had a brainstem stroke had a D FIM for bladder of 1 - 3
Success Rate
Admission comprehension 1–3
88 patients / 50% were successful
If the patient’s comprehension was 1 – 3 and expression was 1 -3
they were more likely to be successful if
in the first 3 hours of SLP they spent time in:
Problem solving Executive function
and/or they:
stayed longer had a hemorrhagic stroke had higher A FIM Cognition and Verbal Expression*
If the patient’s comprehension was 1 – 3 and expression was 1 -3
they were less likely to be successful if
in the first 3 hours of SLP they spent time in:
Reading comprehension
and/or they: were female had a D FIM for bladder of 1 – 3* had a D FIM for bladder of 4 - 5
Success Rate
Admission comprehension 1 – 3 and expression 1 – 3
77 patients / 54.6% were successful
If the patient’s comprehension was 4 – 5 on admission
they were more likely to be successful if
in the first 3 hours of SLP they spent time in:
Problem solving
and/or they: stayed longer had higher A FIM Cognition, Memory*, or Comprehension*
If the patient’s comprehension was 4 – 5 on admission
they were less likely to be successful if
in the first 3 hours of SLP they spent time in:
Auditory comprehension
and/or they: had higher A CSI had a D FIM for bladder of 1 – 3
Success Rate
Admission comprehension 4–5
114 patients / 52% were successful
If the patient’s comprehension was 4 – 5 and expression was 4 -5
they were more likely to be successful if
in the first 3 hours of SLP they spent time in:
Problem solving
and/or they: were white stayed longer had a D FIM for bladder of 6-7 had higher A FIM Cognition or Expression
If the patient’s comprehension was 4 – 5 and expression was 4 -5
they were less likely to be successful if
in the first 3 hours of SLP they spent time in:
Verbal expression
and/or they: had a hemorrhagic stroke stayed longer had a higher A CSI had higher A FIM Motor or Toilet Transfer* were of race: other
Success Rate
Admission comprehension 4– 5 and expression 4 - 5
75 patients / 52% were successful
In Summary, during 1st therapy block
Greater likelihood of success
Less likelihood of success
Complex activities
Mid-level activities
Problem solving, executive functioning
Verbal expression, reading comprehension and written expression
What this means for clinical practice
For patients with low linguistic ability on admission Appear to benefit from high level tasks early in treatment Tasks may not directly correspond to impairment area May need more cueing/assistance to complete tasks initially
Why might this be?
High level tasks involve
critical thinking mental flexibility mental manipulation integration of multiple components of information creativity
Common Threads
Corresponds to results in other therapy disciplines
OT – home management tasks PT – gait and advanced gait
Common Threads
Introducing complex tasks earlier in a length of stay despite the patient’s level of impairment is associated with better outcomes
integration of individual components vs. each component in isolation May require recruitment of more cognitive and linguistic skills that drive functional activity
Error detection, revision/repair, self-regulation
Limitations in data analysis
Limitations of FIM as measurement tool Subjective and objective choices in creating homogenous groups Potential inconsistencies/inaccuracies in recording of treatment interventions Context in which the intervention was implemented was not captured
Future research
Validity studies Interventions Correlation of findings with standardized test scores Initial evaluation time Impact of such a large % of time on swallowing Variation in site practice International comparisons of practice New Zealand CERICE
Acknowledgements
National Institute on Disability & Rehabilitation Research (Grant # H133B990005) RW Brannon, MSPH, project officer RRTC on Medical Rehabilitation Outcomes Boston University with subcontracts to: ICOR (Salt Lake City, UT) & NRH-CHDR (Washington, DC) U.S. Army & Materiel Command (Cooperative Agreement Award # DAMD17-02-2-0032) CR Miles & M Lopez, PhD project officers). NRH Neuroscience Research Center
National Blue Cross-Blue Shield Association Gerben DeJong, PhD
Koen Putnam, PhD
The many contributions of individual clinical sites & investigators